Genome architecture (size, proportion of coding sequences, polycistronic mRNAsŠ) is the consequence of many interacting forces acting over very long time scales. Indeed, selection, mutation biases, drift etc. all act directly or indirectly on the different genome compartments, either triggering their increase or decrease in size or proportion depending on the life style of the organisms (e.g., environment complexity, environment variation, population sizeŠ). Understanding how the intertwining of these pressures leads to particular architecture is difficult for at least three reasons. First, it is not clear how a given change of life-style will affect these different pressures.Second, it is impossible to experimentally separate these forces to observe their individual action since they are all interacting one with the others at the molecular level.Third, they act on time scales that are almost impossible to access experimentally, even with species reproducing rapidly. We propose that in silico experimental evolution approaches, in which simulated species…

Progress in medicine and biology has been so steep that data interpretation has now become an even greater bottleneck than data generation.These data are either structured or unstructured, i.e. found either in databases or in texts; and they tell about biological interactions within species or about evolutionary relationships among them.Our hypothesis is that it is possible to integrate all of these heterogeneous and massive data types coherently via a single network, and that in turn, the topological properties of this networks predict novel relationships and mechanisms that either explain the behavior of an organism as well as novel targets for drugs that modify it.Support for this hypothesis comes from multiple predictions that are experimentally validated, including:a) the identification of protein function and substrates with and without structures; andb) the elucidation of new post-translational modifier-protein target and Protein-Protein  Interaction pairs.Together with other evolutionary tools to identify disease-causing mutations and genes, this integrative network…

24-hour biological clocks are nearly ubiquitous amongst eukaryotes, controlling processes from the human sleep-wake cycle to flowering in plants. Linking the clock’s molecular mechanisms to high-level biological functions, timed to appropriate phases of the day/night cycle and the annual cycle of seasons, exemplifies the challenges of multi-scale systems biology. These “circadian” rhythms share similar properties across all organisms, even in some prokaryotes, yet the key ‘clock proteins’, rhythmic transcriptional regulators, are not conserved across taxa. We are testing the clock circuit in the plant Arabidopsis thaliana, combining molecular and reporter gene timeseries data, with clock mutants and environmental perturbations. The resulting mathematical models of the clock and its output pathways (Biomodels 55, 89, 214, 273, 295, 350, 412, 445) have predicted both particular molecular regulation and operating principles, such as the regulatory flexibility provided by interlocking, negative feedback loops. Testing these hypotheses in the tiny alga Ostreococcus tauri revealed a different strategy to gain flexibility in…

Ammonia oxidation in soil was, until recently, thought to be dominated by autotrophic, chemolithotrophic proteobacteria. These organisms gain energy from oxidation of ammonia to nitrite and perform the first step in nitrification. All cultivated ammonia oxidising bacteria show no or very poor growth in liquid batch culture below pH 7. Paradoxically, nitrification occurs in acid soils at rates that are often faster than those in apparently more favourable neutral and alkaline soils. Understanding acidophilic ammonia oxidation is important globally and economically. Approximately 30% of the world’s soils are acidic (pH<5), including 50% of arable soils, and up to 70% of ammonia-based nitrogen fertiliser is lost through nitrification and leaching or denitrification of nitrate. Potential mechanisms for acidophilic ammonia oxidation include urea hydrolysis, protection in biofilms and aggregates and heterotrophic nitrification, but could also be due to acidophilic or acidotolerant ammonia oxidisers. Assessment of selection for such groups in acid soils was impossible…

The talk will cover results from three as yet unpublished projects. It will start with an introduction into protein sequence searching and to our software package HH-suite for very sensitive remote homology detection. I will show how we can significanlty boost the sensitivity of searches in HH-suite using a homology-enriched profile-HMM database. Second, I will present our new software package MMseqs (Many-against-Many sequence searching) for very fast batch protein sequence searches and clustering of huge protein sequence data sets, such as UniProt or sets of predicted open reading frames from large metagenomics experiments. MMseqs achieves a comparable sensitivity as protein BLAST (blastpgp) at 400x its speed. Third, I present a method for the discovery of regulatory motifs in nucleotide sequences that goes beyond the standard position weight matrix model by learning the dependence of binding affinities on neighboring nucleotides. Importantly, our method automatically learns the dependencies up to the optimum order (i.e.…

.

Genome-wide measurements of transcriptional activity in bacteriaindicate that the transcription of successive genes is stronglycorrelated beyond the scale of operons. I will present an analysis ofhundreds of bacterial genomes to identify supra-operonic segments ofgenes that are proximal in a large number of genomes. The syntenysegments that it yields correspond to genomic units of strongtranscriptional co-expression. Structurally, the segments containoperons with specific relative orientations (co-directional ordivergent) and nucleoid-associated proteins are found to bind at theirboundaries. Functionally, operons inside a same segment are highlyco-expressed even in the absence of regulatory factors at theirpromoter regions. Remote operons along DNA can also be co-expressed iftheir segments share a transcriptional or sigma factor, withoutrequiring these factors to bind directly to the promoters of theoperons. As evidence that these results apply across the bacterialkingdom, I will demonstrate them both in the Gram-negative bacteriumEscherichia coli and in the Gram-positive bacterium Bacillus subtilis.The underlying process that we propose involves only…